If you want to be pedantic then multi-byte string !== UTF-8 support. ;)
Consider the intended purpose of the language and then consider whether the abstraction offered is appropriate. IMO in the case of PHP and UTF-8 it is not.
In my specific case it made my job harder than I would like on 2 projects I used PHP for, which is why I am complaining.
Even your supplied link states they’re not supported at a “low level”, but states nothing about “native”.